WO2018031674A1 - Diagnostic de pathologie amélioré - Google Patents

Diagnostic de pathologie amélioré Download PDF

Info

Publication number
WO2018031674A1
WO2018031674A1 PCT/US2017/046125 US2017046125W WO2018031674A1 WO 2018031674 A1 WO2018031674 A1 WO 2018031674A1 US 2017046125 W US2017046125 W US 2017046125W WO 2018031674 A1 WO2018031674 A1 WO 2018031674A1
Authority
WO
WIPO (PCT)
Prior art keywords
pathology
interest
regions
images
magnified
Prior art date
Application number
PCT/US2017/046125
Other languages
English (en)
Inventor
Joëlle K. Barral
Original Assignee
Verily Life Sciences Llc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Verily Life Sciences Llc filed Critical Verily Life Sciences Llc
Priority to JP2019505009A priority Critical patent/JP6816255B2/ja
Priority to EP17757613.9A priority patent/EP3497620A1/fr
Priority to CN201780049448.4A priority patent/CN109564622B/zh
Publication of WO2018031674A1 publication Critical patent/WO2018031674A1/fr

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/28—Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G06T7/0014—Biomedical image inspection using an image reference approach
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
    • G06V20/698—Matching; Classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/18—Eye characteristics, e.g. of the iris
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/40—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for data related to laboratory analysis, e.g. patient specimen analysis
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
    • G16Z99/00—Subject matter not provided for in other main groups of this subclass
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10056—Microscopic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30024—Cell structures in vitro; Tissue sections in vitro
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/10—Recognition assisted with metadata

Definitions

  • This disclosure relates generally to systems/methods for aiding in pathology.
  • Pathology is a medical discipline related to the study and diagnosis of disease. Most frequently pathology involves the analysis and examination of body-fluid, tissue, and cell samples. As a field of general study and research, pathology relates to four aspects of disease: (1) etiology, (2) pathogenesis, (3) morphologic changes, and (4) consequence of morphologic changes.
  • pathology is a major medical practice that is divided into a number of subdisciplines. In all of these subdiciplines, having a second opinion to check the work of a pathologist is helpful to eliminate erroneous diagnosis. Eliminating diagnostic error in pathology may result in a healthier population and reduce pathologists' exposure to liability.
  • FIG. 1 illustrates a system for pathology diagnosis, in accordance with an embodiment of the disclosure.
  • FIG. 2 illustrates a pathology database to train a machine learning algorithm, in accordance with an embodiment of the disclosure.
  • FIG. 3 illustrates what a pathologist might experience using the system of FIG. 1, in accordance with an embodiment of the disclosure.
  • FIG. 4 is a flow chart illustrating a method of pathology diagnosis, in accordance with several embodiments of the disclosure.
  • the instant disclosure provides a system and method for enhancing pathology diagnosis. More specifically, these systems and methods may be used to aid a pathologist in the diagnosis of disease.
  • a machine learning algorithm may be trained to alert pathologists to regions of interest (e.g., diseased tissue, atypical cells, unusual growth, etc.) in a pathology sample when the machine learning algorithm recognizes these regions in the sample.
  • regions of interest e.g., diseased tissue, atypical cells, unusual growth, etc.
  • Providing a second opinion to pathologists may both increase pathologist efficiency (by allowing the pathologist to spend less time on each slide), and decrease the probability of an inaccurate diagnosis (by alerting the pathologists to regions of interest in the sample).
  • FIG. 1 illustrates a system 100 for pathology diagnosis, in accordance with an embodiment of the disclosure.
  • System 100 includes: microscope 101, digital camera 103, microphone 105, screen 107, processing apparatus 109, network 11 1, storage 123, speaker 125, first mechanical to electrical transducer 113, and second mechanical to electrical transducer 115.
  • the user e.g., a pathologist
  • Microscope 101 is magnifying a pathology sample to form magnified pathology images.
  • the magnified pathology images are recorded with digital camera 103 optically coupled to microscope 101.
  • Digital camera 103 is electrically coupled (wired or wirelessly) to processing apparatus 109 (e.g., a desktop computer, server, etc.) to send the magnified pathology images (still-frames or video) to processing apparatus 109.
  • processing apparatus 109 compares the magnified pathology images to reference pathology images included in a pathology database (contained on remote or local storage 123) to identify one or more regions of interest (e.g., outlined portion 131 on screen 107) in the magnified pathology images. If there is a region of interest in the magnified pathology images, system 100 may alert the user of microscope 101 to the one or more regions of interest.
  • This alert may be an audio alert (e.g., voice from speaker 125 saying "look in the lower left hand corner of the image there may be cancer"), visual alert (e.g., highlighting the region of interest 131 on screen 107), or haptic alert (e.g., vibrating the stage of microscope 101) in response to identifying the one or more regions of interest.
  • an audio alert e.g., voice from speaker 125 saying "look in the lower left hand corner of the image there may be cancer
  • visual alert e.g., highlighting the region of interest 131 on screen 107
  • haptic alert e.g., vibrating the stage of microscope 101
  • a machine learning algorithm (see infra FIG. 2) is used to identify the one or more regions of interest, and is trained to identify the one or more regions of interest using the reference pathology images in the pathology database.
  • the pathology database may include many combinations of reference pathology images, annotations from a pathologist, and grading/diagnosis from a pathologist.
  • image processing and machine learning algorithms may be running so that image data from digital camera 103 is interpreted in real time and guidance can be provided to the user (pathologist) either visually on the monitor or by audio feedback (e.g., "prostate adenocarcinoma Gleason score 6", “slide folded", “image out of focus”, etc.). Headphones can be added to system 100 for convenience, and the conversation can be recorded and archived with respect to the video/images captured by digital camera 103.
  • Reference images in the pathology database may include hundreds of thousands, or even millions, of images of pathology samples that the machine learning algorithm has been trained to recognize as diseased or heathy tissue.
  • the pathology database may also include annotations from a trained pathologist corresponding to the reference pathology images in the pathology database (see infra FIG. 2).
  • the machine learning algorithm may use the annotations from the pathologist in conjunction with the reference pathology images to train itself to identify the one or more regions of interest in the magnified pathology images.
  • the pathology database may include transcribed text that the pathologist spoke when the reference pathology images were recorded.
  • the pathologist recording the images in the database may have said "this portion of the sample looks normal.”
  • the machine learning algorithm may receive this statement from the database as either natural language or transcribed text, and know that the portion of the sample the pathologist is referring to has nothing interesting in it.
  • the machine learning algorithm may note the characteristics of the sample that make it "normal” and recognize these characteristics in future images.
  • the pathologist may have made a statement about the sample being malignant and the machine learning algorithm may note the characteristics of the malignant sample, and apply this knowledge of malignant samples to future magnified pathology images.
  • the pathology database further includes slide position information and magnification information about the reference pathology images in the pathology database.
  • slide position information may have been recorded—this may include the relative coordinates on the sample, the position of the microscope stage, etc.
  • the level of magnification used to see the features in the reference pathology images may also be recorded.
  • the machine learning algorithm may use the slide position information and magnification information to identify the one or more regions of interest in the magnified pathology images.
  • system 100 may identify areas of the pathology sample that are deemed to be insufficiently reviewed by the user, and inform the user of one or more regions of potential interest (e.g., a region of the sample that may appear to system 100 as containing diseased, cancerous, or otherwise interesting tissue/fluid/etc.) in the areas that were deemed to be insufficiently reviewed by the user. For example, if the user of system 100 is examining a pathology sample and inadvertently passes over a region that system 100 recognizes as possibly important, system 100 may alert the user to the presence of the one or more regions of potential interest. Alerting the user may be achieved in the ways discussed above (e.g., via audio, visual, haptic, or other methods).
  • regions of potential interest e.g., a region of the sample that may appear to system 100 as containing diseased, cancerous, or otherwise interesting tissue/fluid/etc.
  • the machine learning algorithm may be trained to identify the areas of the pathology sample that are deemed to be insufficiently reviewed by the user based on the user viewing the areas for a threshold amount of time. In one example, this may involve the machine learning algorithm tracking if a sample was in the user's field of view for a threshold amount of time. For example, if the user of system 100 spends less than one second looking at an area in the pathology sample, system 100 may presume that that area of the sample was passed over.
  • system 100 may be more refined, and use more than just time as the sole metric of if a portion of the sample was insufficiently reviewed. This is because a trained pathologist may simply need to glance at one area of a pathology sample and know immediately it is not of interest. Accordingly, system 100 may determine that a user has/has not viewed specific areas by employing at least one of gaze detection (illustrated as dashed lines from the eyepiece of microscope 101 in FIG. 1), location of the pathology sample (e.g., stage position or relative position on the sample), or level of magnification of microscope 101.
  • gaze detection illustrated as dashed lines from the eyepiece of microscope 101 in FIG. 1
  • location of the pathology sample e.g., stage position or relative position on the sample
  • level of magnification of microscope 101 e.g., level of magnification of microscope 101.
  • system 100 may use gaze detection (e.g., pupal dilation) and alert the pathologist that he/she passed over this portion of the sample.
  • gaze detection e.g., pupal dilation
  • processing apparatus 109 is coupled (wired or wirelessly) to network 111 and/or storage 123.
  • Network 111 and storage 123 may be local or remote, and may or may not be distributed.
  • Storage 123 may include RAM, ROM, hard disks, flash memory, or any other suitable memory system.
  • Network 111 may include the internet or local area network.
  • processing apparatus 109 may be a distributed system on network 111.
  • One skilled in the art will appreciate that there are any number of ways to process/store data in accordance with the teachings of the present disclosure.
  • microphone 105 is electrically coupled to processing apparatus 109 (either wired or wirelessly) to record what the pathologist says. This information may be used to update the pathology database and further teach the machine learning algorithm. Additionally, images captured by digital camera 103 may also be used to update the pathology database along with the microscope stage position (provided by mechanical to electrical transducer 113) and level of magnification (provided by mechanical to electrical transducer 115).
  • system 100 may be used to help pathologists transition to a fully digital microscopy environment, and a mouse is coupled to processing apparatus 109 and/or microscope 101.
  • the mouse may control a motorized stage on microscope 101.
  • the pathologist could choose to either physically move the slide (by turning nobs on the microscope) or the pathologist could move the mouse, and the motorized stage would move the slide to the corresponding location.
  • the motion of the stage could be captured to further inform the machine learning algorithm (e.g., to figure out which parts of the slide have been neglected/passed over).
  • FIG. 2 illustrates a pathology database 201 to train a machine learning algorithm 203, in accordance with an embodiment of the disclosure.
  • Pathology database 201 may be created using a system like system 100 of FIG. 1, and may be stored on storage 123.
  • pathology database 201 includes pathology images (e.g., video or still frames) that were captured by a digital camera (e.g., digital camera 103).
  • the pathology images are indexed with respect to their frame number, recording time, the voice annotation of the pathologist (transcribed), microscope stage position, magnification they were collected at, and location of pathologist gaze.
  • system 100 depicted in FIG. 1 can be used to create a database with any number of dimensions and inputs and is not restricted to those dimensions/inputs depicted here.
  • a digital camera e.g., digital camera 103 optically coupled to a microscope (e.g., microscope 101) may start recording images of pathology samples as a digital video or still frames.
  • Each frame of the video is indexed with respect to its capture time. For example in the depicted embodiment, frame one was captured during the first three microseconds of recording, frame two was captured in the fourth through seventh microseconds of recording, etc.
  • a microphone e.g., microphone 105) may also record the voice annotation of a user of the microscope. The vocal annotations may be converted into text and/or indexed to their respective recording time and video frame.
  • the pathologist while frame one was captured (during the first three microseconds of recording) the pathologist said the word "this"; in subsequent frames the pathologist stated "looks like lymphatic tissue and may be benign.”
  • the system may also record the position of the microscope stage and index it with respect to the recording time and the magnified pathology images.
  • the location of the stage is measured with X, Y coordinates from a (0,0) point which is the lower left hand position of the stage, and the stage movement is measured in microns.
  • stage position should be broadly construed because it is used to identify specific locations on samples, which one skilled in the art will appreciate may be achieved in any number of ways.
  • stage position is determined optically with respect to the dimensions of the slide being imaged, and not with respect to the microscope hardware. As shown, the magnification that a specific frame was viewed with is also recorded with respect to recording time, transcribed text, stage position, and gaze quadrant.
  • the user's gaze may also be indexed with respect to the other dimensions/inputs illustrated and discussed.
  • the gaze of the user/pathologist is measured in quadrants; meaning the image the user sees is subdivided into four sub-images, and the system records which sub-image the user was looking at during the recording time. This may be achieved with hardware/software installed in the microscope, or other external systems, as one skilled in the art will appreciate that there are many different ways to detect gaze.
  • the embodiment depicted here only illustrates very generally where the pathologist/microscope user was looking, in other embodiments the exact coordinates that the user was looking at are recorded.
  • indexing the magnified pathology images and the voice annotation may include tagging the voice annotation of the user to a region of interest in the magnified pathology images.
  • the pathologist's diagnosis of "benign” is associated with stage position coordinates (136, 47) at 40x magnification, and he/she was looking in quadrants 3 and 4. This allows the machine learning algorithm 203 to know exactly where the pathologist was looking when the determination of "benign” was made. Further the machine learning algorithm 203 knows the history of examination (how much of the slide had been examined up to that point).
  • the processing apparatus may further include logic that when executed by the processing apparatus causes the processing apparatus to convert the pathologist's voice annotation to text, and the text is indexed with respect to recording time and the magnified pathology images, among the other dimensions/inputs mentioned and discussed.
  • the pathologist may be able to review the pathology images collected and directly annotate the image to show a region of interest (e.g., circle the cancer cells on the digital image, place a star next to an unknown cell formation, etc.) to make teaching machine learning algorithm 203 easier.
  • a region of interest e.g., circle the cancer cells on the digital image, place a star next to an unknown cell formation, etc.
  • pathology database 201 may be compared to generate a ground truth/argument regarding what is known about the sample/slide. Redundancy of information about a pathology sample may make the diagnosis in pathology database 201 more robust, and provide a larger sample size to train machine learning algorithm 203.
  • Machine learning algorithm 203 may be based on a neural-network type approach or other methods such as association rule learning, deep learning, inductive logic programming, decision tree learning, support vector machines, Bayesian networks, reinforcement learning, clustering, representation learning, similarity and metric learning, sparse dictionary learning, genetic algorithms, or the like. Additionally, machine learning algorithm 203 may include a number of different algorithms running in parallel or at discrete intervals.
  • FIG. 3 illustrates what a pathologist might experience using system 100 of FIG. 1, in accordance with an embodiment of the disclosure.
  • Image 301 is what a user/pathologist might see when looking through microscope 101.
  • image 301 is projected onto screen 107.
  • a speaker e.g., speaker 125 of FIG. 1
  • the regions of interest and potential regions of interest may be highlighted (e.g., outlined) in image 301.
  • the outline may appear on a screen or in the eyepiece of the microscope.
  • highlighting may include laser light shining directly on the sample, or any other way to alert the pathologist examining the sample.
  • a pathologist may be able to start a video chat application either by interacting with the GUI or by calling for help ("Ok microscope, call a breast pathologist").
  • the video feed from the microscope may then be sent to a remote proctor to view.
  • the remote proctor may be able to communicate directly with the local pathologist and system 100. More proctors/specialists can be added as needed. This proctoring capability can also be used for student evaluation.
  • the proctoring capability may be combined with the machine learning algorithm: if the system detects a rare diagnosis, it could suggest calling an expert or even automatically do it. This capability could also be expanded to enable consensus pathology, where each slide is always reviewed by a number of pathologists simultaneously and a diagnosis is made only when a consensus has been achieved (e.g., initially three pathologists are connected, if they disagree a fourth pathologist is added etc. until the desired level of agreement is reached).
  • FIG. 4 is a flow chart illustrating a method 400 of pathology diagnosis, in accordance with several embodiments of the disclosure.
  • the order in which some or all of process blocks 401 - 407 appear in method 400 should not be deemed limiting. Rather, one of ordinary skill in the art having the benefit of the present disclosure will understand that some of method 400 may be executed in a variety of orders not illustrated, or even in parallel.
  • Block 401 illustrates magnifying a pathology sample with a microscope to form magnified pathology images. This may be done manually, by adjusting nobs on a microscope, or may be performed digitally using a computer system to control the microscope. In one embodiment, the microscope could also automatically zoom/move the slide so that the most relevant part of the slide (in order to make the diagnosis) is shown to the pathologist first. This feature could be combined with a slide feeding apparatus to greatly speed up pathologist workflow.
  • Block 403 shows recording the magnified pathology images with a digital camera optically coupled to the microscope.
  • a processing apparatus may receive magnified pathology images from the digital camera either wirelessly or by wired transmission.
  • Block 405 describes comparing the magnified pathology images to reference pathology images included in a pathology database to identify one or more regions of interest in the magnified pathology images.
  • the one or more regions of interest include at least one of diseased portions of the pathology sample, or atypical cells in the pathology sample.
  • the processing apparatus may run a machine learning algorithm to identify the one or more regions of interest, and the machine learning algorithm may be trained to identify regions of interest using the pathology database (see supra FIG. 2).
  • the pathology database may include annotations from a pathologist corresponding to the reference pathology images in the pathology database.
  • the machine learning algorithm may use the annotations from the pathologist in conjunction with the reference pathology images to identify one or more regions of interest in the magnified pathology images.
  • the pathology database further includes magnification information, and location information to identify a position on the reference pathology images.
  • the machine learning algorithm may be trained by, and use the location information and the magnification information about, the reference pathology images to identify the one or more regions of interest in the magnified pathology images.
  • the pathology database further includes a plurality of reference pathology images of the same disease to train the machine learning algorithm (e.g., the database may have many images of melanoma).
  • Block 407 shows alerting a user of the microscope to the one or more regions of interest in the magnified pathology images. This may involve the processing apparatus outputting instructions to a microphone to output speech, highlight a portion of a screen, or moving (e.g., vibrate) a piece of equipment.
  • the machine learning algorithm outputs a diagnosis of a structure in the one or more regions of interest. In another or the same embodiment, the machine learning algorithm may output a confidence interval for the diagnosis, based on the reference pathology images in the pathology database.
  • a tangible non-transitory machine-readable storage medium includes any mechanism that provides (i.e., stores) information in a form accessible by a machine (e.g., a computer, network device, personal digital assistant, manufacturing tool, any device with a set of one or more processors, etc.).
  • a machine-readable storage medium includes recordable/non-recordable media (e.g., read only memory (ROM), random access memory (RAM), magnetic disk storage media, optical storage media, flash memory devices, etc.).

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Theoretical Computer Science (AREA)
  • Public Health (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Epidemiology (AREA)
  • Primary Health Care (AREA)
  • Biomedical Technology (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Pathology (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Library & Information Science (AREA)
  • Ophthalmology & Optometry (AREA)
  • Human Computer Interaction (AREA)
  • Quality & Reliability (AREA)
  • Fuzzy Systems (AREA)
  • Biophysics (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Psychiatry (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Veterinary Medicine (AREA)
  • Physiology (AREA)
  • Signal Processing (AREA)
  • Computing Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)

Abstract

La présente invention concerne un procédé qui comprend l'agrandissement d'un échantillon de pathologie avec un microscope pour former des images de pathologie agrandies, et l'enregistrement des images de pathologie agrandies avec un appareil photo numérique couplé optiquement au microscope. Le procédé comprend également la comparaison des images de pathologie agrandies à des images de pathologie de référence comprises dans une base de données de pathologies pour identifier une ou plusieurs régions d'intérêt dans les images de pathologie agrandies. Un utilisateur du microscope est averti de la présence de la ou des régions d'intérêt dans les images de pathologie agrandies.
PCT/US2017/046125 2016-08-12 2017-08-09 Diagnostic de pathologie amélioré WO2018031674A1 (fr)

Priority Applications (3)

Application Number Priority Date Filing Date Title
JP2019505009A JP6816255B2 (ja) 2016-08-12 2017-08-09 高度な病理診断
EP17757613.9A EP3497620A1 (fr) 2016-08-12 2017-08-09 Diagnostic de pathologie amélioré
CN201780049448.4A CN109564622B (zh) 2016-08-12 2017-08-09 增强病理诊断

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US15/235,541 2016-08-12
US15/235,541 US10025902B2 (en) 2016-08-12 2016-08-12 Enhanced pathology diagnosis

Publications (1)

Publication Number Publication Date
WO2018031674A1 true WO2018031674A1 (fr) 2018-02-15

Family

ID=59700191

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2017/046125 WO2018031674A1 (fr) 2016-08-12 2017-08-09 Diagnostic de pathologie amélioré

Country Status (5)

Country Link
US (3) US10025902B2 (fr)
EP (1) EP3497620A1 (fr)
JP (2) JP6816255B2 (fr)
CN (1) CN109564622B (fr)
WO (1) WO2018031674A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2021515240A (ja) * 2018-04-12 2021-06-17 グーグル エルエルシーＧｏｏｇｌｅ ＬＬＣ 定量的バイオマーカデータのオーバレイを有する病理学用拡張現実顕微鏡
JP2021166062A (ja) * 2018-05-07 2021-10-14 グーグル エルエルシーGoogle LLC 顕微鏡スライド画像のための焦点重み付き機械学習分類器誤り予測
US11594024B2 (en) 2017-06-13 2023-02-28 Google Llc Augmented reality microscope for pathology

Families Citing this family (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6333871B2 (ja) * 2016-02-25 2018-05-30 ファナック株式会社 入力画像から検出した対象物を表示する画像処理装置
US10025902B2 (en) * 2016-08-12 2018-07-17 Verily Life Sciences Llc Enhanced pathology diagnosis
US10733730B2 (en) 2017-06-19 2020-08-04 Viz.ai Inc. Method and system for computer-aided triage
US10346979B2 (en) 2017-06-19 2019-07-09 Viz.ai Inc. Method and system for computer-aided triage
EP3988987A4 (fr) * 2018-09-28 2023-09-20 Evident Corporation Système de microscope, unité de projection et procédé de projection d'image
EP3706128A4 (fr) * 2018-11-09 2021-05-12 Lunit Inc. Procédé de gestion de travail d'annotation et appareil et système le prenant en charge
US10936160B2 (en) * 2019-01-11 2021-03-02 Google Llc System, user interface and method for interactive negative explanation of machine-learning localization models in health care applications
EP3953757A1 (fr) 2019-04-08 2022-02-16 Leica Instruments (Singapore) Pte. Ltd. Microscope à auto-apprentissage
FR3095878B1 (fr) * 2019-05-10 2021-10-08 Univ De Brest Procédé d'analyse automatique d'images pour reconnaître automatiquement au moins une caractéristique rare
JP2021124861A (ja) * 2020-02-04 2021-08-30 ソニーグループ株式会社 解析装置、解析方法、解析プログラム及び診断支援システム
WO2021183955A1 (fr) * 2020-03-13 2021-09-16 PAIGE.AI, Inc. Systèmes et procédés servant à traiter automatiquement des images électroniques à travers des zones
US20230122392A1 (en) * 2020-03-30 2023-04-20 Verily Life Sciences Llc Artificial Intelligence-Based Assistant For Concurrent Review Of Needle Core Prostate Biopsies
WO2021200002A1 (fr) * 2020-03-31 2021-10-07 オリンパス株式会社 Système de microscope, unité de projection et procédé d'assistance à un test d'œuf
WO2021230000A1 (fr) * 2020-05-15 2021-11-18 ソニーグループ株式会社 Dispositif de traitement d'informations, procédé de traitement d'informations et système de traitement d'informations
WO2022044095A1 (fr) * 2020-08-24 2022-03-03 オリンパス株式会社 Dispositif de traitement d'informations, dispositif d'apprentissage, et modèle appris
JP6908806B1 (ja) * 2020-09-16 2021-07-28 ＢｏｎＢｏｎ株式会社 プログラム、情報処理装置、方法
WO2022158908A1 (fr) * 2021-01-22 2022-07-28 주식회사 루닛 Méthode et système d'analyse d'image pathologique
US11694807B2 (en) * 2021-06-17 2023-07-04 Viz.ai Inc. Method and system for computer-aided decision guidance
CN113241184B (zh) * 2021-06-24 2022-07-29 华侨大学 一种儿童肺炎辅助诊断模型及其训练方法
EP4388548A1 (fr) * 2021-08-18 2024-06-26 PAIGE.AI, Inc. Systèmes et procédés destinés au traitement d'images électroniques avec intégration de métadonnées
DE102021121635A1 (de) * 2021-08-20 2023-02-23 Carl Zeiss Microscopy Gmbh Automatisiertes trainieren eines maschinengelernten algorithmus basierend auf der überwachung einer mikroskopiemessung
WO2023128059A1 (fr) * 2021-12-28 2023-07-06 Lunit Inc. Procédé et appareil de pureté de tumeur sur la base d'une image de diapositive pathologique
DE102022116407A1 (de) * 2022-06-30 2024-01-04 Ali Eissing-Al-Mukahal System zur Unterstützung eines Nutzers bei der bildbasierten Erkennung einer Gewebeentartung

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110206283A1 (en) * 2010-02-23 2011-08-25 Pernilla Quarfordt System and method for improved image analysis through gaze data feedback
US20120004514A1 (en) * 2009-03-04 2012-01-05 Atsushi Marugame Diagnostic imaging support device, diagnostic imaging support method, and storage medium
US20120188283A1 (en) * 2011-01-25 2012-07-26 Sony Corporation Image processing device, image processing method and program
WO2014144103A1 (fr) * 2013-03-15 2014-09-18 Sony Corporation Caractérisation des images de pathologie au moyen de l'analyse statistique des réponses de réseau neuronal local
US9373168B2 (en) * 2011-07-13 2016-06-21 Koninklijke Philips N.V. Method for automatically adjusting a focal plane of a digital pathology image

Family Cites Families (39)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2697372B2 (ja) 1991-06-25 1998-01-14 日本電気株式会社 高電力半導体集積回路冷却試験治具
JP3263111B2 (ja) * 1992-01-29 2002-03-04 株式会社東芝 画像保管通信システム及びその端末装置
US6005964A (en) 1996-01-24 1999-12-21 The Board Of Trustees Of The University Of Illinois Automatic machine vision microscope slide inspection system and method
US6396941B1 (en) * 1996-08-23 2002-05-28 Bacus Research Laboratories, Inc. Method and apparatus for internet, intranet, and local viewing of virtual microscope slides
US6441958B1 (en) 2001-03-29 2002-08-27 Chak Sing Richard Yeung Digital imaging microscope
JP2003093380A (ja) * 2001-09-25 2003-04-02 Hitachi Medical Corp 読影支援装置
JP2004135868A (ja) * 2002-10-17 2004-05-13 Fuji Photo Film Co Ltd 異常陰影候補検出処理システム
US7187790B2 (en) * 2002-12-18 2007-03-06 Ge Medical Systems Global Technology Company, Llc Data processing and feedback method and system
US20040122705A1 (en) * 2002-12-18 2004-06-24 Sabol John M. Multilevel integrated medical knowledge base system and method
US7116440B2 (en) 2003-02-28 2006-10-03 Aperio Technologies, Inc. Image processing and analysis framework
US7529394B2 (en) 2003-06-27 2009-05-05 Siemens Medical Solutions Usa, Inc. CAD (computer-aided decision) support for medical imaging using machine learning to adapt CAD process with knowledge collected during routine use of CAD system
JP2005182670A (ja) 2003-12-24 2005-07-07 Icc Computer System:Kk 遠隔画像閲覧システムおよび遠隔画像閲覧用表示方法
EP1787156A1 (fr) 2004-06-11 2007-05-23 Nicholas Etienne Ross Diagnostic automatise de la malaria et d'autres infections
US7958063B2 (en) * 2004-11-11 2011-06-07 Trustees Of Columbia University In The City Of New York Methods and systems for identifying and localizing objects based on features of the objects that are mapped to a vector
US7573439B2 (en) * 2004-11-24 2009-08-11 General Electric Company System and method for significant image selection using visual tracking
CA2595248A1 (fr) * 2005-01-18 2006-07-27 Trestle Corporation Systeme et procede pour la creation d'images de qualite variable d'une diapositive
JP4624841B2 (ja) * 2005-04-13 2011-02-02 オリンパスメディカルシステムズ株式会社 画像処理装置および当該画像処理装置における画像処理方法
CN1891155A (zh) * 2006-05-26 2007-01-10 北京思创贯宇科技开发有限公司 一种基于ct图像的组织成分分析方法
US8189855B2 (en) * 2007-08-31 2012-05-29 Accenture Global Services Limited Planogram extraction based on image processing
JP4558047B2 (ja) 2008-01-23 2010-10-06 オリンパス株式会社 顕微鏡システム、画像生成方法、及びプログラム
JP2010035756A (ja) * 2008-08-04 2010-02-18 Fujifilm Corp 診断支援装置及び診断支援方法
PL2419849T3 (pl) * 2009-04-15 2018-05-30 Koninklijke Philips N.V. Systemy i sposoby wspomagania decyzji klinicznych
JP4676021B2 (ja) * 2009-04-16 2011-04-27 富士フイルム株式会社 診断支援装置、診断支援プログラムおよび診断支援方法
US8995733B2 (en) * 2009-04-28 2015-03-31 Koninklijke Philips N.V. Microdissection method and information processing system
CN102656586B (zh) * 2009-12-18 2016-08-17 皇家飞利浦电子股份有限公司 将采集的图像与对象相关联的方法和系统
DE102010008243B4 (de) * 2010-02-17 2021-02-11 Siemens Healthcare Gmbh Verfahren und Vorrichtung zur Ermittlung der Vaskularität eines sich in einem Körper befindlichen Objektes
CN102253922B (zh) * 2010-05-18 2013-07-03 北京普利生仪器有限公司 远程分析病理切片的方法
US8600143B1 (en) 2010-05-20 2013-12-03 Kla-Tencor Corporation Method and system for hierarchical tissue analysis and classification
US8315812B2 (en) * 2010-08-12 2012-11-20 Heartflow, Inc. Method and system for patient-specific modeling of blood flow
JP5556674B2 (ja) * 2011-01-12 2014-07-23 コニカミノルタ株式会社 医用画像表示装置及びプログラム
EP2845045B1 (fr) 2012-05-02 2023-07-12 Leica Biosystems Imaging, Inc. Mise au point en temps réel en imagerie à balayage linéaire
WO2013175683A1 (fr) * 2012-05-24 2013-11-28 日本電気株式会社 Système d'évaluation des résultats de diagnostics pathologiques, méthode et dispositif d'évaluation associés
US9575304B2 (en) 2012-06-25 2017-02-21 Huron Technologies International Inc. Pathology slide scanners for fluorescence and brightfield imaging and method of operation
RU2640009C2 (ru) * 2012-08-22 2017-12-25 Конинклейке Филипс Н.В. Автоматическое обнаружение и извлечение предшествующих аннотаций, релевантных для визуализирующего исследования, для эффективного просмотра и отчета
FR3024540B1 (fr) * 2014-07-29 2018-02-16 L'air Liquide, Societe Anonyme Pour L'etude Et L'exploitation Des Procedes Georges Claude Debitmetre utilise pour le dosage de l'energie apportee a un procede utilisant un fluide cryogenique
AU2015331579A1 (en) * 2014-10-17 2017-05-25 Cireca Theranostics, Llc Methods and systems for classifying biological samples, including optimization of analyses and use of correlation
JP6514486B2 (ja) 2014-10-29 2019-05-15 ジーイー・メディカル・システムズ・グローバル・テクノロジー・カンパニー・エルエルシー 医用システム及び医用装置並びにプログラム
CN105550651B (zh) * 2015-12-14 2019-12-24 中国科学院深圳先进技术研究院 一种数字病理切片全景图像自动分析方法及系统
US10025902B2 (en) * 2016-08-12 2018-07-17 Verily Life Sciences Llc Enhanced pathology diagnosis

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120004514A1 (en) * 2009-03-04 2012-01-05 Atsushi Marugame Diagnostic imaging support device, diagnostic imaging support method, and storage medium
US20110206283A1 (en) * 2010-02-23 2011-08-25 Pernilla Quarfordt System and method for improved image analysis through gaze data feedback
US20120188283A1 (en) * 2011-01-25 2012-07-26 Sony Corporation Image processing device, image processing method and program
US9373168B2 (en) * 2011-07-13 2016-06-21 Koninklijke Philips N.V. Method for automatically adjusting a focal plane of a digital pathology image
WO2014144103A1 (fr) * 2013-03-15 2014-09-18 Sony Corporation Caractérisation des images de pathologie au moyen de l'analyse statistique des réponses de réseau neuronal local

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
C F NODINE ET AL: "Using eye movements to study visual search and to improve tumor detection", RADIOGRAPHICS, vol. 7, no. 6, 1 November 1987 (1987-11-01), pages 1241 - 1250, XP055416160 *
GURCAN M N ET AL: "Histopathological Image Analysis: A Review", IEEE REVIEWS IN BIOMEDICAL ENGINEERING, IEEE, USA, vol. 2, 1 January 2009 (2009-01-01), pages 147 - 171, XP011507549, ISSN: 1937-3333, DOI: 10.1109/RBME.2009.2034865 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11594024B2 (en) 2017-06-13 2023-02-28 Google Llc Augmented reality microscope for pathology
JP2021515240A (ja) * 2018-04-12 2021-06-17 グーグル エルエルシーＧｏｏｇｌｅ ＬＬＣ 定量的バイオマーカデータのオーバレイを有する病理学用拡張現実顕微鏡
JP2021166062A (ja) * 2018-05-07 2021-10-14 グーグル エルエルシーGoogle LLC 顕微鏡スライド画像のための焦点重み付き機械学習分類器誤り予測
JP7134303B2 (ja) 2018-05-07 2022-09-09 グーグル エルエルシー 顕微鏡スライド画像のための焦点重み付き機械学習分類器誤り予測
US11657487B2 (en) 2018-05-07 2023-05-23 Google Llc Focus-weighted, machine learning disease classifier error prediction for microscope slide images

Also Published As

Publication number Publication date
JP7145195B2 (ja) 2022-09-30
JP2021047911A (ja) 2021-03-25
JP2019533847A (ja) 2019-11-21
US10025902B2 (en) 2018-07-17
JP6816255B2 (ja) 2021-01-20
CN109564622B (zh) 2023-09-01
US10262757B2 (en) 2019-04-16
CN109564622A (zh) 2019-04-02
US11501871B2 (en) 2022-11-15
US20180301217A1 (en) 2018-10-18
EP3497620A1 (fr) 2019-06-19
US20180046759A1 (en) 2018-02-15
US20190198160A1 (en) 2019-06-27

Similar Documents

Publication Publication Date Title
US11501871B2 (en) Enhanced pathology diagnosis
US10545327B2 (en) Pathology data capture
BMid Interpretative phenomenological analysis: Implementing research to influence breastfeeding education
US20220020482A1 (en) Systems and methods for augmented reality-enhanced field services support
JP2012155455A (ja) 画像処理装置および方法、並びにプログラム
US20230177798A1 (en) Relationship modeling and anomaly detection based on video data
US11127494B2 (en) Context-specific vocabulary selection for image reporting
Deka et al. AI-based automated speech therapy tools for persons with speech sound disorder: a systematic literature review
CN108765413B (zh) 用于图像分类的方法、装置及计算机可读介质
Ji et al. Considerations for data acquisition and modeling strategies: Mitosis detection in computational pathology
Gibson et al. Multiple instance learning for behavioral coding
US20210398624A1 (en) Systems and methods for automated intake of patient data
Kocielnik et al. Deep Multimodal Fusion for Surgical Feedback Classification
WO2023107124A1 (fr) Construction d'un système d'article à action pragmatique
US11902650B2 (en) Image capturing apparatus, notification apparatus, notification method, and storage medium
Krappe et al. Automated plasmodia recognition in microscopic images for diagnosis of malaria using convolutional neural networks
Hannes et al. Understanding evidence from research qualitative
Wang et al. Multimodal Depression Detection Network Based on Emotional and Behavioral Features in Conversations
Kuthe et al. INNOVATIVE ALEX AS A HOSPITAL RECEPTIONIST
WO2023220010A1 (fr) Évaluation vidéo et audio basée sur l'intelligence artificielle
Wu et al. Cell nuclei segmentation in divergent images using deep learning and stochastic processing
KR20160095430A (ko) 아이 트래킹 센서를 이용한 독해 지원 장치 및 그 방법
CN118948331A (zh) 一种基于ai大模型的超声检查系统及方法
Hara et al. SIGACCESS
TUYNMAN EMOTION RECOGNITION IN AUDIO SAMPLES CONTAINING SPEECH

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17757613

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019505009

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2017757613

Country of ref document: EP

Effective date: 20190312